Add codespell: config + action (to detect new typos) and make it fix a good number of them

[yarikoptic]

codespell was used occasionally on some files, so not a new tool here. But now it would guard RTD from having typos introduced to begin with.

Note that some fixes could affect (fix) functionality.

There is an odd "variable" syntaxt which I didn't fix since seems would require transition and smells like it is on purpose:

❯ git grep syntaxt -- readthedocs/
readthedocs/projects/migrations/0106_add_addons_config.py:                ("syntaxt", models.CharField(max_length=128)),
readthedocs/projects/models.py:    syntaxt = models.CharField(max_length=128)
readthedocs/projects/tasks/builds.py:        grab them by using glob syntaxt between other files that could be garbage.

or was it intended to be syntax ?

TODOs

• Remove TEMP commit intended to test if pre-commit catches the added malicious typo
• Final decision on fixing or not CHANGELOG.rst
• Pre-commit configuration for codespell common#212 needs to be merged first

---

📚 Documentation previews 📚

• User's documentation (docs): https://docs--11195.org.readthedocs.build/en/11195/

• Developer's documentation (dev): https://dev--11195.org.readthedocs.build/en/11195/

[sentry-io]

🔍 Existing Issues For Review

Your pull request is modifying functions with the following pre-existing issues:

📄 File: readthedocs/api/v2/views/integrations.py

Function	Unhandled Issue
get_closed_external_version_response	TypeError: 'NoneType' object is not subscriptable /api/v2/webhook/{project_slug}/{integration_pk}... Event Count: 1

---

_{Did you find this useful? React with a 👍 or 👎}

[humitos]

Thanks for the PR 👍🏼

I didn't take a full review, but I noted a few issues that I see will generate some conflicts. In particular, those that change Python code and require some manual interaction (e.g. a Django migration to update the help_text)

I'm not opposed to this work, but I find it hard to prioritize by the core team at this point.

[humitos]

This would require a Django migrations.

[yarikoptic]

ok, let me exclude all such lines with help_text *=.*

[yarikoptic]

even better -- while reviewing tuneups needed for pre-commit (e.g. #11201 and from there on) - ralized that it is migrations -- better be excludes altogether.

[yarikoptic]

sorry for the noise -- original is still in readthedocs/builds/models.py . I will add explicit skip for this phrase and TODO

[humitos]

Note we can't skip the whole migrations directory since we are running safemigrations pre commit over it mainly, but also other checks as well

[humitos]

I think there is no need to touch this file. Also, changing the name of the PRs here won't change them in GH and these won't match.

[yarikoptic]

my 1c: IMHO it is worth fixing it as it is "user facing" , whenever PRs with their titles -- developer facing etc.

But if you insist - I can easily skip it. Just one more call ;)

[agjohnson]

We use pre-commit for all other linting checks. I feel if we implement this it should be at pre-commit, not GHA.

[yarikoptic]

Hmm, interesting, my script should've detected presence of pre-commit config... Will have a look wherever get a chance. You do have ci job running the pre-commit I assume? (I don't see report from pre-commit service)

[yarikoptic]

that explains it:

❯ ls -ld .pre-commit-config.yaml
lrwxrwxrwx 1 yoh yoh 29 Mar  7 18:00 .pre-commit-config.yaml -> common/pre-commit-config.yaml
❯ cat .pre-commit-config.yaml
cat: .pre-commit-config.yaml: No such file or directory
❯ git submodule
-4af0fffd2cbeeb40f0a71b875beb99d6dc88a355 common

Let's do the submodules dance now...